Microvascular oxygen tension in the rat mesentery

Aleksander S. Golub, Matthew C. Barker, Roland N. Pittman


Longitudinal Po2 profiles in the microvasculature of the rat mesentery were studied using a novel phosphorescence quenching microscopy technique that minimizes the accumulated photoconsumption of oxygen by the method. Intravascular oxygen tension (Po2, in mmHg) and vessel diameter (d, in μm) were measured in mesenteric microvessels (n = 204) of seven anesthetized rats (275 g). The excitation parameters were as follows: 7 × 7-μm spot size; 410 nm laser; and 100 curves at 11 pulses/s, with pulse parameters of 2-μs duration and 80-pJ/μm2 energy density. The mean Po2 (± SE) was 65.0 ± 1.4 mmHg (n = 78) for arterioles (d = 18.8 ± 0.7 μm), 62.1 ± 2.0 mmHg (n = 38) at the arteriolar end of capillaries (d = 7.8 ± 0.3 μm), and 52.0 ± 1.0 mmHg (n = 88) for venules (d = 22.5 ± 1.0 μm). There was no apparent dependence of Po2 on d in arterioles and venules. There were also no significant deviations in Po2 based on d (bin width, 5 μm) from the general mean for both of these types of vessels. Results indicate that the primary site of oxygen delivery to tissue is located between the smallest arterioles and venules (change of 16.3 mmHg, P = 0.001). In conclusion, oxygen losses from mesenteric arterioles and venules are negligible, indicating low metabolic rates for both the vascular wall and the mesenteric tissue. Capillaries appear to be the primary site of oxygen delivery to the tissue in the mesenteric microcirculation. In light of the present results, previously reported data concerning oxygen consumption in the mesenteric microcirculation can be explained as artifacts of accumulated oxygen consumption due to the application of instrumentation having a large excitation area for Po2 measurements in slow moving and stationary media.

  • microcirculation
  • phosphorescence quenching microscopy
  • oxygen partial pressure profiles
  • oxygen consumption

sequential branching of the arteriolar tree forms microvessels of decreasing diameter, which, in turn, increases the surface area per unit volume available for the diffusion of oxygen to the tissue. In any oxygen-consuming tissue, the rate of intravascular oxygen losses is inversely related to arteriolar vessel diameter, thereby creating an intravascular longitudinal oxygen gradient. In 1970, Duling and Berne (9) originally described findings of substantial intravascular longitudinal gradients in oxygen partial pressure (Po2) from large to small vessels in the arteriolar networks of the hamster and rat. Subsequent studies that employed polarographic and spectroscopic methods confirmed the existence of significant oxygen outflow from arterioles in brain cortex (10, 60), skeletal muscle (23, 2628, 43, 44, 47), hamster skin fold (14, 24, 25, 50), cat retina (4), and ingrowing tumors (7). All of these measurements were made in relatively thick tissue preparations with high metabolic capacities. Furthermore, experimental findings of oxygen losses from arterioles have been analyzed and discussed in several reviews (21, 32, 52, 53, 56, 57), and the findings are further supported by mathematical modeling (36, 39, 40, 42).

The development of longitudinal Po2 gradients in arteriolar networks is, in part, recognized as being a result of the convective transport of oxygen and its subsequent diffusion through the vessel wall to the surrounding tissue. With the assumption that this is the case in tissues having high oxygen consumption (e.g., skeletal muscle), significant oxygen losses would not be expected from mesenteric arterioles where a sparse network of microvessels is surrounded by an extremely thin layer of loose connective tissue (17.4–58.6 μm) covered with mesothelium on either side (1, 13, 63).

Experiments performed in the rat mesentery have indicated the existence of a substantial oxygen gradient in mesenteric arterioles (54, 64, 65). In these studies, the hypothesis was put forth that the vascular tissue has a very high metabolic rate and serves as the primary sink for the oxygen transported by blood. Consequently, arteriolar wall oxygen consumption (V̇o2) governs the magnitude of the longitudinal Po2 gradient instead of the tissue oxygen consumption. Ye et al. (64) concluded that the oxygen consumption rate of arterioles in the rat mesentery is as high as 115 μmol·h−1·g−1 wet wt [about 700 nl O2/(cm3·s)]. In a study by Tsai et al. (54), the oxygen consumption of rat mesenteric vascular tissue was calculated to be 65,000 nl O2/(cm3·s). Similar high values of vessel wall oxygen consumption have been obtained from Po2 measurements made in the rat cremaster muscle and the hamster window chamber preparation (6, 15, 43, 44). For comparison, previous measurements of oxygen consumption by microvascular segments in vitro revealed more moderate values of 60 nl O2/(cm3·s) in hamster mesenteric arterioles <150 μm in diameter and a V̇o2 of 22–60 nl O2/(cm3·s) in cat pial arterioles, 60–300 μm in diameter (20, 30). In addition, the V̇o2 determined in rabbit aortic tissue using an oxygen microelectrode was 80 nl O2/(cm3·s) (3).

The study of longitudinal Po2 gradients in the microvessels of varying tissues can help elucidate the function of microvascular networks as conduits for the supply of oxygen to tissues. Knowledge of longitudinal Po2 gradients, coupled with information concerning local tissue Po2 and oxygen consumption, provides a more complete analysis of microvascular function. Yet, upon review of the literature, there are disparities among reported values for similar tissues. This review has led us to hypothesize the following for the current study. First, if the arteriolar wall has a V̇o2 that is hundreds of times higher than any known tissue V̇o2, then longitudinal Po2 gradients are mainly determined by the arteriolar wall V̇o2 that is equally high in all organs. Therefore, second, the longitudinal gradient in mesenteric arterioles is as steep as those reported for organs with the highest oxidative capacity. For this purpose, distributions of intravascular Po2 in mesenteric microvessels of different types and diameters were measured with a phosphorescence quenching microscopy (PQM) method. To evaluate the transmural Po2 drop in arterioles, intravascular Po2 was compared with perivascular Po2 measured with the scanning PQM technique under the identical physiological conditions (16).


Animal preparation.

Seven female Sprague-Dawley rats (275 ± 6 g; Harlan, Indianapolis, IN) were used in this study. Animals were initially anesthetized with a combination of ketamine and acepromazine (75 and 2.5 mg/kg ip). Surgical preparation and measurements were conducted under a continuous intravenous infusion of Saffan (0.1 mg·kg−1·min−1; Schering-Plough Animal Health, Welwyn Garden City, UK) following catheterization of the right jugular vein. Upon conclusion of the experimentation, animals were euthanized with the administration of Euthasol (0.4 ml/kg iv; Delmarva, Midlothian, VA). All procedures and protocols were approved by the Institutional Animal Care and Use Committee of Virginia Commonwealth University.

Initially, the trachea was cannulated with polyethylene tubing (PE-240) to ensure a patent airway. Subsequent catheterization of the left carotid artery with PE-50 tubing allowed for monitoring of mean arterial pressure (CyQ 103/301, Cyber Sense, Nicholasville, KY) and arterial blood gas sampling. Measured arterial blood gas variables included Po2, Pco2, and pH (ABL 705, Radiometer America, Westlake, OH).

After catheterization the animal was placed on a thermostatic animal platform against an organ pedestal (17). The heated (37°C) sapphire window of the pedestal was painted with a yellow fluorescent marker (for visualization of the excitation spot) and covered with Saran Wrap film (Dow Chemical, Indianapolis, IN) fixed with a neoprene ring around the pedestal. A loop of small intestine with mesentery was exteriorized via a cauterized midline abdominal incision and placed around the pedestal window on the supporting neoprene foam ring. A Saran film circle the size of the pedestal window was placed on the mesentery. A larger sheet of Saran film was then placed on top, and its edges were wrapped around the intestine and attached to the neoprene ring with pushpins. An objective-mounted pressure bag made of Saran film was filled with a thin layer of water for immersion of the ×40 objective and pressurized to 8–12 mmHg (16). To reduce the noise and exclude unnecessary excitation of the phosphorescent probe, the experiments were carried out in a dark room and imaged with filtered transillumination (OG-570 orange glass filter; Edmund Optics, Barrington, NJ).

Po2 measurements in microvessels.

A detailed description of our PQM method was published recently (16). Several features in our design of the PQM method make it suitable for microscopic measurements of Po2 in microvessels. To improve the localization of the signal and to avoid unnecessary irradiation of the tissue and vasculature, a small aperture in the epi-illumination light path of the microscope is used. The adjustable aperture is parfocal with the objective lens and determines the size and shape of the excitation spot. The detection arm of the optical path is maximally opened for collection of emitted light so that detection efficiency is determined mainly by the numerical aperture of the objective lens. Such a design provides high sensitivity that allows a good signal-to-noise ratio to be obtained at relatively low flash energy. A short, rectangular excitation pulse (2 μs) and short delay (2 μs) in signal acquisition improve the quality of the signal necessary for exponential analysis of heterogeneous decays.

For Po2 measurements in flowing blood, we used a stationary, 7 × 7-μm excitation spot placed in the centerline of vessels with a light pulse duration of 2 μs and an energy density of 81 pJ/ μm2. Each Po2 measurement was based on an analysis of 100 averaged decays collected at an 11-Hz excitation pulse rate. Vessels chosen for measurement were selected randomly and sorted into bins based on the following diameter ranges: 10–15, 15–20, 20–25, and 25–35 μm. For venules, the largest diameter interval was 25–50 μm instead of 25–35 μm. Capillaries were defined as the smallest microvessels (below 10 μm id) connecting diverging and converging bifurcations. The number of measurements collected in each bin was approximately equal but not proportional to the natural occurrence of the microvessels of the sampled diameter. The internal diameter of vessels (d) was measured at the site of excitation with a magnification of 0.35 μm/mm (×40 objective). Po2 and diameter measurements in arterioles and venules were made away from bifurcations (at least 5 × d), and capillary Po2 and diameter measurements were made toward their arteriolar end.

A solution of albumin-bound Pd-meso-tetra-(4-carboxyphenyl)porphyrin (Pd-MTCPP), prepared as previously described (16), was slowly injected intravenously (∼0.5 ml over 5 s). The volume of the probe for injection was calculated individually for each animal to create an initial probe concentration of 0.5 mg/ml in the plasma. It was recently reported that a detectable concentration of the probe did not appear in the extravascular space of skeletal muscle, even after 35 min of perfusion (35). However, mesenteric microvessels are more “leaky” than those in skeletal muscle (i.e., have lower plasma protein reflection coefficients). Consequently, tests were performed to investigate probe extravasation from the vasculature in the mesentery (see results). We determined that the maximal duration of the experiment was limited by probe extravasation from the microvessels and the consequent contribution of extravascular phosphorescence to the intravascular signal. Therefore, we started the Po2 measurements in arterioles, venules, and capillaries 5 min after probe injection and concluded them 30 min later. By the end of this period, the lymphatic microvessels had accumulated a detectable amount of the probe, and their Po2 values were measured.

Data analysis.

Phosphorescence decay curves with a high signal-to-noise ratio were used for calculations of intravascular Po2. A nonlinear fitting procedure of the phosphorescence decay curves was based on the Rectangular model of Po2 distribution (16, 18). In addition, several tests were conducted using the phosphorescence signal to establish validity of signal quality and degree of oxygen consumption by the method itself (i.e., the use of a stationary excitation spot).

Statistical calculations and linear and nonlinear fitting procedures were made using the Origin 7.0 program (OriginLab, Northampton, MA). All data presented in tables are means ± SE (n = number of measurements), and the differences of two means were analyzed using an unpaired or paired t-test, as appropriate. The significance of differences between compared mean values was at the P = 0.05 level, unless otherwise specified.



Measurements of phosphorescence amplitude (proportional to the amount of probe in the excitation region) were used to verify both homogeneity of the probe distribution in the plasma 5 min after injection and the small contribution of tissue phosphorescence to the signal after a measurement period of 35 min. Figure 1 shows a representative recording of the phosphorescence amplitude in an arteriole during the first 30 s after probe injection and then at time points collected at about 5 min of postinfusion. This test demonstrates that 5 min following probe injection, the plasma concentration of Pd-MTCPP in microvessels is stable enough to begin Po2 measurements.

Fig. 1.

The amplitude of the phosphorescence signal (in V), proportional to the probe concentration in the plasma, as recorded in a 21-μm arteriole over time. The plasma concentration of the probe was variable during the first 30 s and was found to be stable in a 10-s period (after break on abscissa) recorded 5 min after the probe was injected. Measurements were made at a rate of 10 per s and smoothed with a 5-point adjacent averaging procedure.

To estimate any contribution of extravasated probe to the overall phosphorescence signal, we measured the signal amplitude over the centerline of a small arteriole and then in the tissue at a point 57 μm away at the end of a 35-min interval. Signal amplitudes were 5.94 and 2.68 V, respectively. Using the micrograph of a rat mesentery cross-section from Fox and Wayland (13), we found that a vessel occupies close to 50% of the total tissue thickness. Assuming that the tissue layer above a vessel centerline comprises ∼25% of the overall thickness, we concluded that the extravascular signal component was no higher than 11%. The tissue below the vessel does not significantly contribute to the signal because of its minimal excitation due to the short penetration depth of the violet light in the blood. This estimation indicates that the permeability of albumin in capillaries and venules in the mesentery is higher than in skeletal muscle (35), and, therefore, intra- and extravascular Po2 measurements in connective tissue should be limited to a 30-min period.

Oxygen consumption by the method was tested in blood samples taken from the animals at the end of the measurement period and sealed in rectangular glass microslides (optical path, 100 μm; and temperature, 37°C). The aim of this test was to estimate the maximal Po2 drop caused by a single excitation flash in the sampled volume. This Po2 drop was too small to be measured directly, so a continuous recording of the Po2 at an 11-Hz flash rate was performed in stationary blood samples starting from the first flash. The sequence of mean Po2 values for 20 groups of 11 measurements in each group is presented in Fig. 2. The time course of the Po2 decrease in the excitation volume was approximated by a monoexponential curve, which was then differentiated and plotted against Po2 values to obtain the oxygen consumption rates, as shown in Figs. 2 and 3. This oxygen consumption curve was then extrapolated to the moment of the first flash. The maximal value of the Po2 decrement from the plot in Fig. 3 demonstrates that the depression in oxygen tension caused by the first flash was 0.15 mmHg/flash. Another approach to the evaluation of the consumption artifact was based on the analysis of the initial slope of the curve (Fig. 2), resulting in a higher value of 0.28 mmHg/flash. This estimate (0.15–0.28 mmHg/flash) is in agreement with the results of a previously reported test (33). The single flash photoconsumption artifact is an inevitable part of the measuring procedure, yet, the systematic error remained below 0.3 mmHg, a satisfactory accuracy limit for most physiological studies. Since our technique excluded multiple excitation of the same volume in flowing blood in microvessels, the single flash oxygen consumption artifact was not accumulated during the time interval of Po2 measurement.

Fig. 2.

A continuous recording of Po2 over time in a stationary blood sample contained in a glass microslide at a flash rate of 11 pulses/s. The data were averaged for each second and fitted using exponential analysis to estimate the maximal oxygen consumption per flash by the methods (see Fig. 3).

Fig. 3.

The exponential curve (shown in Fig. 2) was differentiated and plotted vs. Po2 to find the maximal oxygen consumption value that corresponds to the first light pulse (0.15 mmHg/flash). The magnitude of the consumption artifact determined by the Po2 drop between two initial points results in a somewhat higher estimate (0.28 mmHg/flash).

Po2 distribution in microvessels.

Intravascular Po2 was measured in mesenteric vascular networks located in the transparent connective tissue windows. A total of 78 arterioles with internal diameters smaller than 35 μm, 88 venules smaller than 50 μm, and 38 capillaries with diameters of ∼8 μm were measured at random within the animal group (n = 7; systemic variables are shown in Table 1). Measurements from a given vessel were placed into bins based on vessel classification (i.e., arteriole, venule, or capillary) and diameter. The Po2 distribution in the microvessels of the rat mesentery is presented in Fig. 4.

Fig. 4.

The distribution of Po2 in the microvascular network of the rat mesentery. Diameters (d) of arterioles are shown as negative values and diameters of venules are positive for the convenience of presentation. Data are presented as means ± SE; numbers of vessels in a group are shown above the symbols.

View this table:
Table 1.

Systemic animal variables

No significant differences or trends were found between mean Po2 values in arterioles of different diameters. The largest and smallest arterioles (d = 28.4 ± 0.6 and 12.0 ± 0.3 μm, respectively) had the same Po2 (65.7 ± 2.4 and 66.6 ± 2.9 mmHg, respectively). The Po2 in venules was also the same for all diameters ranging from 33.0 ± 1.2 to 13.1 ± 0.3 μm. The corresponding values of Po2 for these venular diameter groups were 53.9 ± 2.6 and 50.3 ± 3.0 mmHg, respectively.

The maximal difference in Po2 (16.3 mmHg, P = 0.001) was found between the smallest arterioles (d = 12.0 ± 0.3 μm; Po2 = 66.6 ± 2.9 mmHg; n = 25) and venules (d = 13.1 ± 0.3 μm; Po2 = 50.3 ± 3.0 mmHg; n = 20). The difference of the mean Po2 values between the entire arteriolar (d = 18.8 ± 0.7 μm; Po2 = 65.0 ± 1.4 mmHg; n = 78) and venular (d = 22.5 ± 1.0 μm; Po2 = 52.2 ± 1.4 mmHg; n = 88) sets of vessels was 12.8 mmHg (P = 0.001). The Po2 of capillaries (d = 7.8 ± 0.3 μm; Po2 = 62.1 ± 2.0 mmHg; n = 38) was biased toward the arteriolar level and significantly higher than the total venular Po2 (P = 0.01).

Lymphatic Po2 measurements were carried out in relatively large lymphatic vessels located in radial vascular bundles. Lymphatic Po2 (76.3 ± 3.4 mmHg) was measured in 15 vessels. Two vessels were half covered with adipose tissue, so the internal diameter (76.2 ± 8.1 μm) was measured in only 13 of the 15 vessels.


Results summary.

Based on in vivo testing, we determined that the period of Po2 measurements in mesenteric microvessels should be limited to 35 min after the probe was injected to ensure the intravascular localization of the phosphorescence signal. In addition, in vitro testing also determined that the oxygen consumption artifact was lower than 0.3 mmHg/flash and did not create a significant systematic error of Po2 measurements in flowing blood. In the present study, the mean arteriolar Po2 values in the different diameter vessels exhibited no significant difference. Similarly, the mean venule Po2 values did not change significantly among venular diameter sizes. Based on these results, there is no evidence to support the supposition of a substantial longitudinal Po2 gradient in either the arteriolar or venular components of the mesenteric microvascular network. The capillary bed was the only microvascular site where a longitudinal Po2 drop was revealed. The difference between the mean arteriolar and venular Po2 values was determined to be 12.8 mmHg.

Longitudinal Po2 gradients.

Substantial longitudinal Po2 gradients and oxygen losses in arterioles have been confirmed for many thick tissues having high metabolic rates as determined by polarographic (4, 9, 28, 60), spectroscopic (26, 27, 37, 48), and phosphorescence quenching (7, 23, 24, 43, 45, 54) techniques. Po2 distributions determined by spectroscopic and fluorescence-quenching methods in the microvessels of the rat mesentery vary from what has been found in thicker tissues. A spectroscopic study (48) found that oxygen release is the greatest in capillaries under normal conditions, yet oxygen losses became apparent in both the arterioles and venules when the mesentery was superfused with a deoxygenated solution containing dithionite. The fluorescence-quenching measurements (22, 63), performed in a mesentery preparation isolated from ambient air, found a negligible longitudinal gradient in arterioles and venules under conditions where the animal was ventilated with room air. A substantial longitudinal Po2 gradient became detectable only when animals were switched to breathing 100% oxygen. Conversely, Tsai et al. (54) reported, using a PQM technique to measure Po2 in mesenteric arterioles, significant longitudinal oxygen losses (expressed in saturation units of 2.4%/100 μm). The present study, also performed with a PQM technique, supports the finding that in the mesentery, under normal conditions, longitudinal Po2 gradients in arterioles and venules are not substantial and that the capillaries are the main site of oxygen exchange in the mesenteric vasculature. It is probable that contiguous longitudinal Po2 measurements following the branching orders in an arteriolar network would possibly reveal a longitudinal oxygen gradient, but it was not evident in the statistical approach used in the present work.

The absence of an apparent longitudinal Po2 gradient in the mesenteric microcirculation is directly due to low tissue and vessel wall V̇o2. The rat mesentery represents a thin (17–58 μm thick) sheet of loose connective tissue, without any parenchymal cells, that is covered by thin mesothelium on either side (1, 13, 63). Tissue oxygen consumption in the rat mesentery has recently been determined with a scanning PQM method. Golub et al. (16) reported that the oxygen consumption of the mesenteric tissue was below 60–68 nl O2/(cm3·s), which is typical for connective tissue (8, 61), yet 25% of that previously reported (54). In the interstitial Po2 study (16), a probe was topically applied to the mesentery before measurements to localize the phosphorescence signal to the extravascular compartment. Conversely, in the present study, the phosphorescence signal was localized to the intravascular space via intravenous probe infusion and a 30-min measurement period limitation. This approach allows the data from the two separate experiments to be compared without having to take into account the relative contribution of the phosphorescence signal resulting from both compartments. The intravascular Po2 in arterioles (65.0 ± 1.4 mmHg; n = 78) measured in the present work was very close to the tissue Po2 values above (62.7 ± 2.0 mmHg; n = 84) and beside (62.2 ± 2.1 mmHg; n = 84) arterioles (16). Thus the data from both experiments are consistent with the assertion that the transmural Po2 difference (<3 mmHg) is statistically and physiologically insignificant in the mesentery.

Recent advances in the development of phosphorescent probes have allowed simultaneous measurements of intra- and extravascular Po2 distributions in a large-sampled volume using two oxygen probes localized in the two different compartments (62). These experiments revealed a small Po2 difference between the interstitium and the plasma in resting skeletal muscle. The small transmural Po2 drop, found in the dual probe experiments and in our results obtained with separate probe applications, indicates that oxygen consumption of the vascular wall in both the mesentery and skeletal muscle at rest is minimal.

Oxygen saturation.

An estimation of oxygen saturation, based on a rat blood oxygen saturation curve (41), showed that the blood in the smallest arterioles was saturated at 84% (Po2 = 67 mmHg) and blood in the smallest venules was saturated at 70% (Po2 = 50 mmHg). The saturation drop between the smallest arterioles and venules was 14%. These data are identical to the results of independent measurements made in the same classes of vessels with Raman spectroscopy of hemoglobin (51). An explanation for the uncharacteristically high oxygen saturation in collecting venules of the mesentery is twofold. First, as previously reported by Golub et al. (16), the oxygen consumption of mesenteric tissue is low. Secondly, the relatively high oxygen saturation of the blood leaving the mesenteric vasculature is certainly related to the position of the mesentery between systemic arteries and the hepatic portal system, which provides the primary oxygen supply to hepatic tissue. This anatomical relation makes the vascular network of the mesentery, especially its venous part, functionally different from the vasculature of most other tissues.

Lymphatic vessel Po2.

In addition to making intravascular arteriolar and venular Po2 measurements, we also measured Po2 in lymphatic vessels located among transit bundles of mesenteric vessels partially covered with adipose tissue. Lymphatic vessels have been described as having their own blood-carrying microvascular network (19), and this can explain why the Po2 in lymph could be even higher than that of small arterioles inside the mesenteric windows. Reported results of polarographic Po2 measurements in mesenteric lymph in dogs corroborate our finding of high Po2 in the large mesenteric lymphatic vessels (11).

Velocity limitations for the measurement of Po2 in moving media.

The parameters of oxygen transport obtained in the present work, in conjunction with our previous measurements of periarteriolar Po2 (16), provide a distinctly different perspective on microvascular oxygen transport compared with previously reported results for the rat mesentery and similar tissue types by other groups of investigators (5, 6, 15, 19, 25, 54, 55). We propose that the drastic differences among the results can be explained primarily by the use of varying measurement methodologies, embodied by the different designs of the measuring instruments. As such, we have scrutinized the source of these discrepancies in connection with the employed PQM instrument design and the methods of data interpretation in the present work.

The principal disparity between the instruments is the size of the excitation and detection areas. The instrument employed by Torres Filho and Intaglietta (49) and Tsai et al. (54) combined a large excitation area (LEA; diameter, 140 μm) with a small detection area in the excitation area center (5 × 20 μm). The instrument used in the present study had a small excitation area (SEA; 7 × 7 μm) combined with a large detection area (diameter > 360 μm). These intrinsic features (i.e., either LEA or SEA parameters) determine instrument performance concerning Po2 measurements under conditions of varying velocities of the sampled fluid volumes.

The oxygen probe is a photosensitizer that causes oxygen consumption to occur in response to the excitation light (5, 12, 29, 33, 59). The amount of consumption caused by each excitation pulse depends on the probe concentration, the energy density of the excitation light, the concentration of dissolved oxygen, and the concentration of organic molecules available for photooxidation. Since the probe concentration and light intensity are kept constant during multiple flash excitation and since we assume the concentration of organic molecules available for photooxidation to be relatively constant, the oxygen consumption by a single flash caused by photooxidation (or photoconsumption) is proportional mainly to the concentration of oxygen (5, 29).

The probe concentration and energy density are selected empirically to keep the decrease of Po2 per flash below 1 mmHg (5, 33). Because of these limitations and the microscopic size of the sampled volume, the signal-to-noise ratio of an individual decay curve is usually too low for the purpose of exponential analysis. To improve the signal quality, multiple excitation pulses (at a frequency, F, during the time period, T) are used to accumulate an average curve that is sufficient for accurate computation of Po2 (46, 49, 66). When velocity, V, of the sampled fluid in a microvessel is high enough, each excitation flash excites a new volume of moving fluid, so that the average curve will contain the same small error of Po2 caused by a single flash. At a velocity lower than some critical level (V < Vcr), the sampled volume undergoes multiple flash excitations. In that case, the Po2 decrease in the sampled volume will accumulate and become dependent on the number of light pulses that a particular volume receives. Accordingly, the average decay curve will be affected by this accumulated artifact of photoconsumption to a degree inversely related to the velocity. Thus the slowest sampled volume will have the largest underestimate of Po2.

The magnitude of the consumption artifact per second in multiple excitation measurements depends on the combination of the excitation area radius, R, and flash rate, F. In the LEA instrument used by Tsai et al. (54), the critical velocity is determined by the flowing sample volume from the edge of the excitation area to the central detection region for the time between two flashes: Vcr = F·R = 30 Hz × 70 μm = 2,100 μm/s. In our SEA instrument, a new sampled volume enters the excitation area between two flashes, when Vcr = F·2R = 11 Hz × 7 μm = 77 μm/s (where 2R represents the excited volume of radius, R, shifted to a distance of two times the radius). The linear velocity of all mesenteric vessels exceeds 77 μm/s, which makes our instrument applicable to Po2 measurements in moving blood without limitations. For measurements in stationary media, a small excitation spot can be scanned along a region of interest to avoid multiple excitation of the same sampled volume (16). Additional advantages of our SEA design include the absence of unnecessary illumination of other tissues in the microscopic field and a wide aperture for phosphorescence light collection, which improves the overall sensitivity of the instrument.

In contrast, the LEA instrument employed by Tsai et al. (54) to measure intravascular Po2 in microvessels of the hamster skinfold chamber, where all venules and most arterioles have mean velocities below Vcr = 2,100 μm/s (25), results in an accumulated photoconsumption and therefore an underestimate of the intravascular Po2. In the rat, mesentery terminal arterioles, capillaries, and most venules have velocities below 2,100 μm/s (38), which also affect measurements in this tissue preparation made with the LEA technique. In addition, the penetration depth of the excitation light in blood is short (∼50% attenuation at 7 μm for 410 nm excitation) due to light absorbance by hemoglobin. Thus most of the phosphorescence signal in microvessels is emitted by a peripheral plasma layer with a linear velocity that is lower than the mean velocity. Therefore, an accumulated consumption artifact could occur, even if the mean velocity in a vessel is higher than the Vcr defined for the measurement system. The photoconsumption in plasma cannot be effectively compensated for by oxygen release from red blood cells between flashes because of the time limitation of this release process (33 ms at F = 30 Hz) (58). In the stationary sampled volume, accumulated consumption by the method is maximal because the photoconsumption that occurs in the tissue cannot be compensated for by diffusion from the surrounding volume due to the large size of the depletion region (16).

Longitudinal Po2 gradients and wall oxygen consumption in arterioles.

It is typical in microvascular networks for smaller vessels to have lower linear blood velocities (38, 67). Application of a LEA instrument to vessels with velocities below Vcr = 2,100 μm/s creates a progressive Po2 underestimation from large arterioles to capillaries due to an increase in the accumulated oxygen consumption with the decrease in velocity. In addition, a LEA instrument depresses Po2 in the perivascular tissue (16), thereby creating a greater oxygen sink, which adds to the diffusional oxygen losses from arterioles and amplifies the apparent longitudinal Po2 gradient. Previous workers have concluded that this apparent longitudinal Po2 gradient found in precapillary arterioles is due to high oxygen losses from the arterioles, which led them to further conclude that arterioles are the main site of oxygen supply to the tissue (5254, 56).

In the study of Tsai et al. (54), the longitudinal oxygen saturation drop in mesenteric arterioles with an average diameter of 23 μm was estimated to be 2.4%/100 μm at a Po2 of 43 mmHg. This steep longitudinal Po2 gradient was in good agreement with the low Po2 found in perivascular tissue measured with the LEA technique (54). The transmural Po2 drop, measured in the arterioles of rat mesentery in this work, was 18 mmHg.

We have previously demonstrated that the photoconsumption caused by the instrument employed by Tsai et al. (54) was underestimated due to a flaw in the test (16). Recent assessments of photoconsumption in this instrument resulted in values of 0.84–0.39 μM/flash (about 0.6–0.3 mmHg/flash) (5). These values are in general agreement with the photoconsumption rate found in the present work and in our previous tests involving xenon flash lamp excitation (33). With the LEA method employed by Tsai et al. (54), intravascular Po2 measured in flowing blood is decreased proportionally to the number of flashes received by the sampled volume on its way from the edge to the center of the excitation area where the detection region is located (i.e., the site of Po2 measurement). For example, for a blood velocity as low as 210 μm/s, the number of flashes received by a sampled volume is 10 (for R = 70 μm, F = 30 Hz) and the accumulated consumption is proportional to the product of 0.3 mmHg/flash and 10 flashes. The Po2 decrease in perivascular tissue is determined by the total number of flashes for 3 s of illumination (90 flashes). Thus the accumulated photoconsumption of oxygen in the tissue is much higher than that in a vessel. The impact of photoconsumption on measurements in flowing and stationary media produces a large apparent transmural Po2 drop that could be interpreted as being caused by an extremely high rate of metabolic activity in the arteriolar wall (54). The resultant assertion of the existence of a large transmural Po2 drop found with the LEA method has formed the foundation for the concept that the vascular wall is a primary oxygen sink and that the extremely high vascular wall metabolism serves to protect tissue from the high O2 content of the blood (5254, 56).

The high respiration rate [3.9 ml O2/(cm3·min)] found in the vessel wall (54) gives strong support to the “hot pipes” hypothesis (65) proposed to explain nonshivering thermogenesis as derived from vascular wall heat production. With the use of the energy conversion factors from the work of Ye et al. (65) (4.83 kcal/l O2), the energy production by mesenteric arterioles could be presented as 1,300 W/l of vascular tissue, a value high enough for “hot pipes” thermogenesis.

The specific oxygen consumption of rats was reported to be 28.1 ml O2/(kg·min) at rest and 81.3 ml O2/(kg·min) at exhaustive exercise (2). Thus the range of oxygen consumption of a rat with body mass of 300 g is 8.4–24.4 ml O2/min. The estimated vascular mass (3.4% of body mass; see Ref. 31) for this rat is 10.2 g, which makes the total vascular tissue oxygen consumption 3.9 times 10.2 = 39.8 ml O2/min. In this estimation, the vascular wall oxygen consumption is five times the total animal consumption at rest and 1.6 times higher than the maximum possible value. Even excluding the venous contribution of the vascular oxygen consumption cannot change the fact reported by Tsai et al. that vascular wall oxygen demand significantly exceeds the maximal possible supply in this organism.

Previous estimates of the respiration rate of the arterial wall tissue made with microelectrodes were found to be quite modest [in nl O2/(cm3·s)]: 60 (20), 22–63 (30), 80 and 155 (3). Comparisons of perivascular Po2 with our data, made with the scanning PQM method (16), coupled with the intravascular Po2 measurements obtained in the present work, demonstrate that no significant transmural Po2 drop exists. Extrapolation of the periarteriolar Po2 profile in mesenteric tissue (16) to the blood/wall interface indicated that intravascular Po2 values fit into this curve. This means that the wall oxygen consumption is not significantly different from that of the tissue [60 − 68 nl O2/(cm3·s)]. Recent reports of simultaneous Po2 measurements in blood and tissue in skeletal muscle, made with two different phosphorescent probes, also suggested that there was no significant transmural Po2 difference (62).

Measurements made by the LEA technique in arterioles with arrested flow (6) and in lymphatic microvessels (19) were also affected by the photoconsumption artifacts typical for stationary measurements, as discussed above. The rate of oxygen loss from motionless blood in arterioles was about 10 times faster than the rate of oxygen loss when using a SEA technique (34). Lymphatic microvessels studied with a LEA technique were characterized as having the lowest Po2 levels (22–25 mmHg) in the microvasculature (19, 53), whereas the measurements of lymphatic vessels in the present study using a SEA technique were close to arteriolar values.


The superior accuracy of the SEA technique (compared to the LEA method) in Po2 measurements of slow-moving and stationary fluids gives strong support to our findings of insignificant longitudinal Po2 gradients and a small transmural Po2 drop in arterioles of the rat mesentery. The largest Po2 gradient in the mesenteric microvasculature was found in capillaries. In addition, lymph in the largest lymphatic vessels of the mesentery appeared to be well oxygenated. Therefore, we have concluded that oxygen consumption of the arteriolar wall and the surrounding tissue in the mesentery is much lower than has previously been determined by the use of LEA methods. The reported findings with LEA instruments of large longitudinal and transmural Po2 gradients in mesenteric arterioles and low Po2 in lymph are, consequently, an intrinsic artifact of accumulated oxygen consumption by the method due to multiple excitation of the same sampled volume in flowing blood and stationary tissue.


This study was supported by National Heart, Lung, and Blood Institute Grants HL-18292 and HL-79087 and America Heart Association, Mid-Atlantic Affiliate, Grant AHA-0655449U.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract